A smart trolley obstacle avoidance system based on image processing

Abstract

The application provides a kind of intelligent car obstacle avoidance system based on image processing, compared with prior art, the application also includes three-wheeled car, set on the three-wheeled car for image shooting of the situation in front of three-wheeled car camera component, the image information photographed by camera component is preprocessed to identify the road area and background area in image preprocessing module, the further analysis and identification of the bend and bend direction of the road in front of three-wheeled car identification module and the further control of three-wheeled car driving based on the identification of identification module control module.The application realizes the identification of the bend situation of the road in front of three-wheeled car, and adjusts the vertical spacing of the front wheels of the vehicle through the intelligent control system, optimizes the stability and safety of the vehicle driving in the bend, effectively controls the body tilt of the system, reduces the risk of rollover, enhances the stability of the outside in the turning process, improves the safety and efficiency of the three-wheeled car in the driving process.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of trolley obstacle avoidance systems, and in particular to an intelligent trolley obstacle avoidance system based on image processing. BACKGROUND

[0002] Intelligent three-wheeled trolleys are small mobile platforms that combine automatic driving, Internet of Things, and artificial intelligence technologies, and are commonly used in fields such as logistics distribution, security patrol, home service, and education and scientific research. With the continuous progress of intelligent and automated technologies, these trolleys have achieved autonomous navigation, real-time obstacle avoidance, and remote control functions through components such as sensors, GPS, control systems, and communication modules.

[0003] The experimental team has long been engaged in the research and development of intelligent trolley technology, and has conducted a large number of relevant experiments. Through extensive research, the team found that existing technologies such as CN110244718B, CN110182205B, CN114253273B, and CN113885532B have certain limitations. For example, the existing technology discloses a control system for an intelligent obstacle-avoiding trolley, which includes a trolley sensing module that collects sensing data of the trolley during its movement based on a pre-set ultrasonic sensing device and uploads the data to an analysis module. The analysis module receives the sensing data and analyzes it, and then transmits the analysis results to a control module. The control module reads the analysis results and generates target control instructions, and controls the trolley to move normally based on the instructions. This system can accurately determine the location of obstacles on the trolley's path and plan the trolley's path, improving the accuracy of the trolley's obstacle avoidance and the safety of its movement.

[0004] To address the problem of inaccurate road recognition during trolley movement in the field, the present application was developed. SUMMARY

[0005] The present application aims to address the shortcomings in the field by providing an intelligent trolley obstacle avoidance system based on image processing.

[0006] To overcome the shortcomings of the prior art, the present application adopts the following technical solutions:

[0007] The application discloses a kind of based on image processing intelligence's trolley obstacle avoidance system, the trolley obstacle avoidance system includes tricycle, set in the tricycle for the image shooting of tricycle front situation camera component, the image information that camera component is photographed is preprocessed to identify the road area and background area in image pre-processing module, the further analysis identification of the curve of tricycle front road and curve direction identification module and the further control of tricycle travel based on the identification of identification module Control module.

[0008] Further, the tricycle includes a vehicle body, a rear wheel arranged at the bottom of the vehicle body and corresponding to the rear end of the vehicle body in the driving direction of the vehicle body, a steering controller arranged on the vehicle body for controlling the rotation direction of the rear wheel, a motor driver for driving the rear wheel to rotate, two front wheels arranged at the bottom of the vehicle body and corresponding to the front end of the vehicle body in the driving direction of the vehicle body, and two electric telescopic control rods respectively for controlling the distance between the front wheels and the vehicle body.

[0009] Further, two open grooves are arranged on the bottom wall of the vehicle body, each open groove includes a groove opening on the bottom wall and a groove cavity corresponding to the groove opening and recessed upward, the groove cavity includes a cavity wall arranged on five sides, and one side of the cavity wall opposite to the groove opening is the top cavity wall of the groove cavity.

[0010] Further, the distance between each front wheel and the vehicle body is adjusted by the extension and contraction of the electric telescopic control rods, one of the two electric telescopic control rods is used to control the distance between the left front wheel and the vehicle body, the other electric telescopic control rod is used to control the distance between the right front wheel and the vehicle body, the electric telescopic control rod for controlling the distance between the left front wheel and the vehicle body is the left electric telescopic control rod, and the electric telescopic control rod for controlling the distance between the right front wheel and the vehicle body is the right electric telescopic control rod.

[0011] The top of the left electric telescopic control rod is fixed to the top cavity wall of one of the open grooves, and the bottom is fixed to the left front wheel, and the top of the right electric telescopic control rod is fixed to the top cavity wall of the other open groove, and the bottom is fixed to the right front wheel.

[0012] Further, the distance between the right front wheel and the vehicle body is kept unchanged, the vertical distance between the wheel shaft of the left front wheel and the vehicle body is adjusted by the extension of the left electric telescopic control rod, thereby increasing the vertical distance between the left front wheel and the vehicle body, tilting the vehicle body to the right, increasing the grip force of the right front wheel, and providing more stability when the tricycle drives on a right curve.

[0013] Further, the distance between the left front wheel is kept unchanged, and the vertical distance between the right front wheel axle and the vehicle body is adjusted by the elongation operation of the right electric telescopic control rod, thereby increasing the vertical distance between the right front wheel and the vehicle body, making the vehicle body tilt to the left, increasing the grip of the left front wheel, and providing more stability when the three-wheeled vehicle drives on the left curve.

[0014] Further, the control module is electrically connected with the two electric telescopic control rods to control the telescopic operation of the two electric telescopic control rods.

[0015] When the recognition module determines that the road ahead of the three-wheeled vehicle is a left curve, the control module controls the elongation operation of the right electric telescopic control rod.

[0016] When the recognition module determines that the road ahead of the three-wheeled vehicle is a right curve, the control module controls the elongation operation of the left electric telescopic control rod.

[0017] The beneficial effects obtained by the present application are:

[0018] 1. The present application adjusts the vertical distance between the front wheels on both sides of the vehicle body and the vehicle body to control the tilting of the vehicle body, which effectively resists the centrifugal force generated on the outer side during curve driving, provides more lateral support by increasing the ground pressure of the inner side tire, thereby increasing stability and reducing the risk of rollover, to enhance the stability of the three-wheeled vehicle under certain steering conditions, so that the three-wheeled vehicle exhibits better maneuverability and safety under various road turning requirements.

[0019] 2. The present application realizes the identification of the curve condition of the road ahead of the three-wheeled vehicle, and adjusts the vertical distance of the front wheels of the vehicle through the intelligent control system, optimizes the stability and safety of the vehicle during curve driving, so that the system not only accurately analyzes the boundary movement in the fused image to realize real-time curve recognition, but also dynamically adjusts the height configuration of the vehicle to adapt to different driving environments, automatically adjusts the distance between the front wheels and the vehicle body, effectively controls the tilting of the vehicle body, reduces the risk of rollover, enhances the stability on the outer side during turning, and improves the safety and efficiency of the three-wheeled vehicle during driving.

[0020] 3. The present application adjusts the vertical distance between the front wheels on both sides of the vehicle body and the vehicle body to control the tilting of the vehicle body, which effectively resists the centrifugal force generated on the outer side during curve driving, provides more lateral support by increasing the ground pressure of the inner side tire, thereby increasing stability and reducing the risk of rollover, to enhance the stability of the three-wheeled vehicle under certain steering conditions, so that the three-wheeled vehicle exhibits better maneuverability and safety under various road turning requirements. BRIEF DESCRIPTION OF DRAWINGS

[0021] The application can be further understood from the following description in conjunction with the drawings. The components in the drawings are not necessarily drawn to scale, but emphasis is instead placed upon illustrating the principles of the embodiments. In different views, the same reference numerals designate corresponding parts.

[0022] Figure 1 A modular schematic diagram of the image processing intelligent-based car obstacle avoidance system of the present application.

[0023] Figure 2 A working flow schematic diagram of the preprocessing module of the present application.

[0024] Figure 3 A working flow schematic diagram of the recognition module of the present application. DETAILED DESCRIPTION

[0025] In order to make the objectives, technical solutions and advantages of the present application clearer and more comprehensible, the present application is further described in detail below in conjunction with its embodiments; it is pointed out that the specific embodiments described herein are only used to explain the present application and are not used to limit the present application. Other systems, methods and / or features of the present embodiments will become apparent to those skilled in the art after reading the following detailed description. And the terms used to describe the positional relationship in the drawings are only used for illustrative description and cannot be understood as limiting the present patent; for those skilled in the art, the specific meanings of the above terms can be understood according to the specific circumstances.

[0026] Embodiment one: in conjunction with the attached Figure 1 , the attached Figure 2 and the attached Figure 3 , the present embodiment constructs an image processing intelligent-based car obstacle avoidance system, the car obstacle avoidance system comprising a three-wheeled car, a camera assembly arranged on the three-wheeled car for image shooting of the situation in front of the three-wheeled car, a preprocessing module for preprocessing the image information shot by the camera assembly to identify the road area and the background area in the image, a recognition module for further analyzing and identifying the curve of the road in front of the three-wheeled car and the direction of the curve, and a control module for further controlling the driving of the three-wheeled car based on the recognition of the recognition module,

[0027] The three-wheeled car comprises a car body, a rear wheel arranged at the bottom of the car body and corresponding to the rear end of the car body in the driving direction of the car body, a steering controller arranged on the car body for controlling the turning direction of the rear wheel, a motor driver for driving the rear wheel to turn, two front wheels arranged at the bottom of the car body and corresponding to the front end of the car body in the driving direction of the car body, and two electric telescopic control rods respectively for controlling the distance between the front wheels and the car body,

[0028] The bottom wall of the vehicle body is provided with two open grooves, each of which comprises a groove opening on the bottom wall and a groove cavity recessed upward from the groove opening, the groove cavity comprises a cavity wall arranged on five sides, and one side of the cavity wall opposite to the groove opening is the top cavity wall of the groove cavity,

[0029] The front wheels are universal wheels of the prior art, and the front wheel located on the left side of the vehicle body in the driving direction is the left front wheel, and the front wheel located on the right side of the vehicle body in the driving direction is the right front wheel,

[0030] The distance between each front wheel and the vehicle body is adjusted by the extension and retraction of the electric telescopic control rods, one of the two electric telescopic control rods is used to control the distance between the left front wheel and the vehicle body, the other electric telescopic control rod is used to control the distance between the right front wheel and the vehicle body, and the electric telescopic control rod used to control the distance between the left front wheel and the vehicle body is the left electric telescopic control rod, and the electric telescopic control rod used to control the distance between the right front wheel and the vehicle body is the right electric telescopic control rod,

[0031] Specifically, the top of the left electric telescopic control rod is fixed to the top cavity wall of one of the two open grooves, and the bottom is fixed to the left front wheel, and the top of the right electric telescopic control rod is fixed to the top cavity wall of the other open groove, and the bottom is fixed to the right front wheel,

[0032] The distance between the left front wheel and the vehicle body is adjusted by the extension of the left electric telescopic control rod, and the vertical distance between the left front wheel and the vehicle body is increased, so that the vehicle body is tilted to the right, the ground adhesion of the left front wheel is increased, and more stability is provided when the three-wheeled vehicle drives on a right curve, and the possibility of rollover is reduced,

[0033] The distance between the right front wheel and the vehicle body is adjusted by the extension of the right electric telescopic control rod, and the vertical distance between the right front wheel and the vehicle body is increased, so that the vehicle body is tilted to the left, the ground adhesion of the right front wheel is increased, and more stability is provided when the three-wheeled vehicle drives on a left curve, and the possibility of rollover is reduced,

[0034] The control module and the two electric telescopic control rods are electrically connected to control the extension and retraction of the two electric telescopic control rods, the control module controls the extension of the right electric telescopic control rod when the identification module determines that the road ahead of the three-wheeled vehicle is a left curve, and the control module controls the extension of the left electric telescopic control rod when the identification module determines that the road ahead of the three-wheeled vehicle is a right curve,

[0035] The application controls the tilting of the vehicle body by adjusting the vertical distance between the front wheels on both sides of the vehicle body and the vehicle body, and the tilting of the vehicle body effectively resists the centrifugal force generated on the outer side during cornering, and provides more lateral support by increasing the ground pressure of the inner tires, thereby increasing stability and reducing the risk of rollover, to enhance the stability of the three-wheeled vehicle under certain steering conditions, so that the three-wheeled vehicle exhibits better maneuverability and safety under various road turning requirements.

[0036] Embodiment two: combined with the accompanying Figure 1 , the accompanying Figure 2 and the accompanying Figure 3 , in addition to containing the contents of the above embodiments, the pre-processing module realizes the following operation steps:

[0037] S101: receiving the image shot by the camera assembly,

[0038] S102: applying a Gaussian filter to remove image noise, using a histogram equalization method to enhance image contrast, so that the visual difference between road and non-road areas is more obvious, and obtaining a pre-processed image,

[0039] S103: analyzing the texture features of the pre-processed image through the gray level co-occurrence matrix, the texture features including contrast, uniformity, and homogeneity indicators, based on the differences in the texture features of the road and non-road areas, and then distinguishing the road and non-road areas, the identified road and non-road areas are represented as a first binary image, wherein the pixels of the road area in the first binary image are marked as 1, and the pixels of the non-road area are marked as 0,

[0040] S104: based on the gray level threshold of the road and non-road areas in the image, the image is segmented into road and non-road areas, and the identified road and non-road areas can be represented as a second binary image, wherein the pixels of the road area in the second binary image are marked as 1, and the pixels of the non-road area are marked as 0,

[0041] S105: fusion processing of the first binary image and the second binary image to obtain a fusion image with fusion marks:

[0042] When both images are marked as road at the same position, the same position is determined as road in the fusion image;

[0043] When both images are background at the same position, the same position is determined as background in the fusion image;

[0044] When the first binary image and the second binary image are marked differently at the same position, taking the pixel point at the same position as a target pixel point in the pre-processed image, the gray level co-occurrence confidence C GLCM and the gray level threshold confidence CThre A calculation comparison is made:

[0045] If C GLCM > C Thre , the same position in the fusion image adopts the marking result determined by the first binary image,

[0046] If C GLCM ≤ C Thre , the same position in the fusion image adopts the marking result determined by the second binary image;

[0047] S106: output the fusion image after fusion processing;

[0048] Wherein, the first binary image of S103 and the second value image of S104 are realized by using OpenCV and skimage provided by the existing related experience threshold, function and method of the neighborhood technical personnel, and will not be repeated here,

[0049] Wherein, the calculation steps of the gray level co-occurrence confidence C GLCM of the target pixel point in the preprocessed image in the above step S105 are as follows:

[0050] S201: select a 7x7 square window centered on the target pixel point of the preprocessed image,

[0051] S202: obtain the contrast value Con, uniformity value Uni and homogeneity value Hom in the 7x7 square window through gray level co-occurrence matrix analysis calculation,

[0052] Wherein, the contrast value Con, uniformity value Uni and homogeneity value Hom are calculated and obtained by using the existing related functions in the OpenCV and skimage library of the existing Python language, and will not be repeated here,

[0053] S203: obtain the gray level co-occurrence confidence C GLCM :

[0054] C GLCM = w1x Con + w2x Uni + w3x Hom,

[0055] Wherein, w1, w2 and w3 are gray level co-occurrence weight factors, and w1+w2+w3=1;

[0056] Wherein, the calculation steps of the gray level threshold confidence C Thre of the target pixel point in the preprocessed image in the above step S105 are as follows:

[0057] S301: Apply Sobel operator, apply 3x3 Sobel kernel to each pixel point in 7x7 square window, and calculate the horizontal gradient value Gx and vertical gradient value Gy of each pixel point in 7x7 square window respectively,

[0058] S302: Calculate the gradient amplitude Gxy of the i-th pixel point in the 7x7 square window as follows:

[0059]

[0060] wherein Gx(i) is the horizontal gradient value of the i-th pixel point in the 7x7 square window, Gy(i) is the vertical gradient value of the i-th pixel point in the 7x7 square window, i = 1, 2, 3…49,

[0061] S303: Calculate the average value Ag of the gradient amplitudes of all pixels in the 7x7 square window:

[0062]

[0063] S304: Move the 3x3 sub-window in the 7x7 window, the 3x3 sub-window can be moved from the top left corner of the window to the right and down in the 7x7 window, and finally cover the entire area of the 7x7 window,

[0064] There are 25 3x3 sub-windows in different positions in the 7x7 window, and the standard deviation SS of each 3x3 sub-window in the 7x7 square window is calculated:

[0065]

[0066] wherein Gra(m) is the gray value of the m-th pixel point in the 3x3 sub-window, and u is the average gray value of all pixel points in the 3x3 sub-window,

[0067] S305: Calculate the average value ALS of the standard deviations of all sub-windows:

[0068]

[0069] wherein SS(j) is the standard deviation of the j-th sub-window in the 7x7 window,

[0070] S306: Obtain threshold segmentation confidence C Thre :

[0071] C Thre =Gxy×w4+w5×ALS,

[0072] w4 and w5 are gray threshold weight factors, and w4+w5=1;

[0073] The application significantly improves the distinguishing ability of the automatic driving car in a complex environment for the front road and the surrounding environment by innovatively combining the edge detection of the Sobel operator and the texture analysis of the gray level co-occurrence matrix, not only realizes the efficient fusion of edge and texture information at the algorithm level, but also optimizes the data processing flow by finely adjusting the weight of local and global features, and enhances the accuracy and robustness of image recognition.

[0074] Embodiment three: in combination with the attached Figure 1 , attached Figure 2 and attached Figure 3 , in addition to containing the contents of the above embodiments, the recognition module realizes the following steps:

[0075] S401: scanning the fusion image, wherein each pixel position in the fusion image recognition is represented by a coordinate position (x, y), wherein x represents the horizontal coordinate value and y represents the vertical coordinate value, and the coordinate origin (0, 0) of the fusion image image is at the lower left corner of the image, then the bottom line of the fusion image is the first line, and the position of the kth line of the fusion image corresponds to the position of the k+1th line of the car driving area in front of the area,

[0076] S402: determining the position of the left boundary and the right boundary of each row in the fusion image, and the left boundary L(k) of the kth row and the right boundary L(k) of the kth row in the fusion image are determined in the following manner:

[0077] L(k) = min[x|M(x, y) = 1],

[0078] R(k) = max[x|M(x, y) = 1],

[0079] min[x|M(x, y) = 1] represents the x corresponding value of the first pixel point coordinate position marked as 1 from left to right in the kth row of the fusion image,

[0080] max[x|M(x, y) = 1] represents the x corresponding value of the last pixel point coordinate position marked as 1 from left to right in the kth row of the fusion image,

[0081] S403: obtaining the left boundary movement amount Delta L(k) in the fusion image:

[0082] Delta L(k) = L(k+1) - L(k),

[0083] S404: obtaining the right boundary movement amount Delta R(k) in the fusion image:

[0084] Delta R(k) = R(k+1) - R(k),

[0085] S405: Calculate the left boundary moving average value LShift of the preset number M of rows in the fusion image:

[0086]

[0087] S406: Calculate the right boundary moving average value RShift of the preset number M of rows in the fusion image:

[0088]

[0089] S407: If LShift>left curve threshold Tleft, then judge that the road in front of the three-wheeled vehicle is a left curve,

[0090] If RShift>right curve threshold Tright exists, then judge that the road in front of the three-wheeled vehicle is a right curve,

[0091] Tleft and Tright are respectively determined by the initial threshold value evaluated by the neighborhood technology and a large amount of historical data by statistical analysis, and further optimized and dynamically adjusted by support vector machine, decision tree and / or deep learning model after collecting boundary moving data under left curve and right curve driving conditions through field test to obtain Tleft for judging left curve and Tright for judging right curve,

[0092] In the embodiment, the person skilled in the art can refer to the above-mentioned value examples and adjust them according to the actual situation. In summary, the specific value range and trend of Tleft, Tright, W1, W2, W3, W4, W5 will be determined by the designer or operator of the processing system according to the actual electromagnetic measurement processing needs and system response, which will not be repeated here,

[0093] The present application realizes the recognition of the curve condition of the road in front of the three-wheeled vehicle, and adjusts the vertical distance of the front wheels of the vehicle through the intelligent control system, optimizes the stability and safety of the vehicle driving in the curve, so that the system can not only accurately analyze the boundary movement in the fusion image and realize real-time curve recognition, but also dynamically adjust the height configuration of the vehicle to adapt to different driving environments, automatically adjust the distance between the front wheels and the vehicle body, effectively control the inclination of the vehicle body, reduce the risk of rollover, enhance the stability of the outside during turning, and improve the safety and efficiency of the three-wheeled vehicle during driving.

[0094] While the application has been described with reference to various embodiments, it will be understood that many modifications and variations of the present application are possible. It is therefore understood that within the scope of the application, that the application can be practiced otherwise than as specifically described. That is, the methods, systems and devices discussed above are examples. Various configurations can omit, substitute, or add various procedures or components as appropriate. For instance, the methods described can be performed in an order different from that described, and / or various steps can be added, omitted or combined. Also, features described with respect to certain configurations can be combined in other configurations, for example, features described with respect to one configuration can be combined with features described with respect to a different configuration. Also, control and signal lines can be conveyed by a variety of means known in the art and the example embodiment, such as a bus, a signal path, a wired connection, wireless connection, etc. Additionally, a plurality of different creative synthesis mechanisms can be employed. Also, the word "comprising" does not exclude the presence of elements or steps other than those listed and the word "a" or "an" preceding the name of an element does not exclude the presence of a plurality of such elements or steps. It is further understood that devices of the present application can optionally include one or more elements, features or steps described herein.

Claims

1. An image processing intelligence-based trolley obstacle avoidance system, characterized in that, The trolley obstacle avoidance system comprises a three-wheeled trolley, a camera assembly arranged on the three-wheeled trolley and used for image shooting of a situation in front of the three-wheeled trolley, a preprocessing module used for preprocessing image information shot by the camera assembly to identify a road region and a background region in the image, an identification module used for further analyzing and identifying a curve of a road in front of the three-wheeled trolley and a curve direction, and a control module used for further controlling trolley driving based on identification of the identification module. The three-wheeled trolley comprises a trolley body, a rear wheel arranged at a bottom of the trolley body and corresponding to a rear end of the trolley body in a driving direction of the trolley body, a steering controller arranged on the trolley body and used for controlling a rotation direction of the rear wheel, a motor driver used for driving the rear wheel to rotate, two front wheels arranged at the bottom of the trolley body and corresponding to front ends of the trolley body in the driving direction of the trolley body, and two electric telescopic control rods respectively used for controlling a distance between the front wheels and the trolley body. The identification module implements the following steps: S401: scanning a fusion image output by the preprocessing module, wherein each pixel position in the fusion image identification is represented by a coordinate position (x, y), wherein x represents a horizontal coordinate value and y represents a vertical coordinate value, and a coordinate origin (0, 0) of the fusion image is located at a lower left corner of the image, a first row in the fusion image is a bottom row, and a position of a kth row in the fusion image is corresponding to a position of a k+1th row which is a region more in front of the automobile driving, S402: determining positions of a left boundary and a right boundary of each row in the fusion image, and the left boundary L(k) of the kth row and the right boundary R(k) of the kth row in the fusion image are determined in the following manner: L(k)=min[x|M(x,y)=1], R(k)=max[x|M(x,y)=1], min[x|M(x,y)=1] represents a corresponding value of x of a first pixel point coordinate position marked as 1 from left to right in the kth row in the fusion image, max[x|M(x,y)=1] represents a corresponding value of x of a last pixel point coordinate position marked as 1 from left to right in the kth row in the fusion image, S403: obtaining a left boundary moving amount ΔL(k) in the fusion image: ΔL(k)=L(k+1)−L(k), S404: obtaining a right boundary moving amount ΔR(k) in the fusion image: ΔR(k)=R(k+1)−R(k), S405: calculating a left boundary moving average value LShift of a preset row number M in the fusion image: LShift = 0x20 , S406: calculating a right boundary moving average value RShift of the preset row number M in the fusion image: RShift = RShift + 1 , S407: when LShift>Tleft, it is determined that a road in front of the three-wheeled trolley is a left curve, and if RShift>Tright, it is determined that the road in front of the three-wheeled trolley is a right curve.

2. The trolley obstacle avoidance system of claim 1, wherein, The bottom wall of the vehicle body is provided with two open grooves, each of which comprises a groove opening on the bottom wall and a groove cavity recessed upward from the groove opening, the groove cavity comprises a cavity wall arranged around five sides, and one side of the cavity wall opposite to the groove opening is the top cavity wall of the groove cavity.

3. The trolley obstacle avoidance system of claim 2, wherein, The distance between the right front wheel and the vehicle body is kept unchanged, and the vertical distance between the axle of the left front wheel and the vehicle body is adjusted by the elongation operation of the left electric telescopic control rod, thereby increasing the vertical distance between the left front wheel and the vehicle body, so that the vehicle body is inclined to the right, the grip of the right front wheel is increased, and more stability is provided when the three-wheeled trolley drives on a right curve.

4. The trolley barrier system of claim 3, wherein, The distance between the right front wheel and the vehicle body is kept unchanged, and the vertical distance between the axle of the right front wheel and the vehicle body is adjusted by the elongation operation of the right electric telescopic control rod, thereby increasing the vertical distance between the right front wheel and the vehicle body, so that the vehicle body is inclined to the left, the grip of the left front wheel is increased, and more stability is provided when the three-wheeled trolley drives on a left curve.

5. The trolley barrier system of claim 4, wherein, The control module is electrically connected with the two electric telescopic control rods to control the telescopic operation of the two electric telescopic control rods. When the identification module determines that the road ahead of the three-wheeled trolley is a left curve, the control module controls the elongation operation of the right electric telescopic control rod. When the identification module determines that the road ahead of the three-wheeled trolley is a right curve, the control module controls the elongation operation of the left electric telescopic control rod.

6. The trolley barrier system of claim 4, wherein, The distance between each front wheel and the vehicle body is adjusted by the telescopic adjustment of the electric telescopic control rod, one of the two electric telescopic control rods is used to control the distance between the left front wheel and the vehicle body, the other of the two electric telescopic control rods is used to control the distance between the right front wheel and the vehicle body, and the electric telescopic control rod used to control the distance between the left front wheel and the vehicle body is the left electric telescopic control rod, and the electric telescopic control rod used to control the distance between the right front wheel and the vehicle body is the right electric telescopic control rod. The top of the left electric telescopic control rod is fixed to the top cavity wall of one of the two open grooves, and the bottom is fixed to the left front wheel, and the top of the right electric telescopic control rod is fixed to the top cavity wall of the other open groove, and the bottom is fixed to the right front wheel. The preprocessing module realizes the following operation steps: S101: receiving an image captured by a camera assembly, S102: applying a Gaussian filter to remove image noise, using a histogram equalization method to enhance image contrast, so that the visual difference between road and non-road areas is more obvious, and obtaining a preprocessed image, S103: analyzing the texture features of the preprocessed image by a gray level co-occurrence matrix, the texture features including contrast, uniformity and homogeneity indicators, based on the differences in texture features between road and non-road areas, and distinguishing road and non-road areas, the identified road and non-road areas are represented as a first binary image, wherein the pixels of the road area in the first binary image are marked as 1, and the pixels of the non-road area are marked as 0, S104: based on the gray threshold of the road and non-road area in the image, the image is segmented into road and non-road area, and the identified road and non-road area can be expressed as a second binary image, wherein the pixels of the road area in the second binary image are marked as 1, and the pixels of the non-road area are marked as 0, S105: fusion processing is performed on the first binary image and the second binary image to obtain a fusion image with fusion marking: When both images are marked as road at the same position, the same position is determined as road in the fusion image; When both images are background at the same position, the same position is determined as background in the fusion image; When the first binary image and the second binary image are different in the same position mark, taking the pixel point at the same position as a target pixel point in the preprocessed image, calculating and comparing the gray level co-occurrence confidence C GLCM and the gray level threshold confidence C Thre of the target pixel point in the preprocessed image. If C GLCM > C Thre , the same position in the fusion image adopts the marking result determined by the first binary image, If C GLCM ≤ C Thre , the same position in the fusion image then adopts the marking result determined by the second binary image; S106: output the fusion image after fusion processing.

Images